Ribosomal Proteins PO, P1, and P2 Are Phosphorylated by Casein Kinase I1 at Their Conserved Carboxyl Termini*

A potential casein kinase I1 (CK 11) recognition site is located within the conserved carboxyl (COOH) terminus of the ribosomal P (phospho) proteins PO, P1, and P2. To determine whether the COOH termini of the P proteins are physiological substrates for CK 11, we studied the phosphorylation of the P proteins in vitro and in intact cells. The results show that the addition of exogenous purified CK I1 and ATP to intact ribosomes in vitro resulted in the relatively selective phosphorylation of all three P proteins. A synthetic peptide corresponding to the COOH-terminal22 amino acids of P2 (C-22) was also phosphorylated by CK I1 with a K,,, of 13.4 p ~ . An endogenous ribosome-asso- ciated, CK 11-like enzyme also phosphorylated the P proteins relatively selectively in the presence of 10 mM Mg2+ and ATP. The endogenous kinase was inhibited by heparin, utilized either ATP or GTP as a phosphate donor, and phosphorylated casein. and the (3-22 peptide inhibited the phosphorylation of the P proteins by the endogenous kinase, providing further evidence for its CK 11-like properties and for localization of the The peptides were first separated by high voltage electrophoresis on cellulose thin layer plates at 800 V for 45 min in 1% NH,HCOa buffer, pH 8.9 (32), followed by ascending thin layer chromatography in the second dimension using pyridine/ n-butyl alcohol/acetic acid/H,O = 60:75:15:60. Tryptic phosphopeptides were located by autoradiography.

A potential casein kinase I1 (CK 11) recognition site is located within the conserved carboxyl (COOH) terminus of the ribosomal P (phospho) proteins PO, P1, and P2. To determine whether the COOH termini of the P proteins are physiological substrates for CK 11, we studied the phosphorylation of the P proteins in vitro and in intact cells. The results show that the addition of exogenous purified CK I1 and ATP to intact ribosomes in vitro resulted in the relatively selective phosphorylation of all three P proteins. A synthetic peptide corresponding to the COOH-terminal22 amino acids of P 2 (C-22) was also phosphorylated by CK I1 with a K,,, of 13.4 p~. An endogenous ribosome-associated, CK 11-like enzyme also phosphorylated the P proteins relatively selectively in the presence of 10 mM Mg2+ and ATP. The endogenous kinase was inhibited by heparin, utilized either ATP or GTP as a phosphate donor, and phosphorylated casein.
A  peptide (Arg-Arg-Arg-Glu-Glu-Glu-Thr-Glu-Glu-Glu) and the (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22) peptide inhibited the phosphorylation of the P proteins by the endogenous kinase, providing further evidence for its CK 11-like properties and for localization of the CK I1 phosphorylation site to the COOH termini of the P proteins. Tryptic phosphopeptide maps of P1 and P2 phosphorylated by exogenous CK I1 and the endogenous ribosome-bound kinase were virtually identical. These phosphopeptides comigrated with the tryptic digest of C-22 and with the tryptic phosphopeptides derived from P1 and P2 isolated from intact cells metabolically labeled with [32P]orthophosphate in uivo. These studies demonstrate that exogenous CK I1 and a ribosome-bound, CK 11-like enzyme phosphorylate the ribosomal P proteins in vitro and localize the target site for phosphorylation to the COOH terminus. The incorporation of phosphate into the same target site in intact cells indicates that the P proteins are in uivo substrates of CK 11.
The eukaryotic P (phospho) proteins PO, P1, and P2 are acidic ribosomal proteins which are located within the 60 S ribosomal subunit (1,2). Cross-linking studies and their known homology to the Escherichia coli ribosomal protein L7/L12 (3), suggest that one P1 homodimer and one P2 homodimer are attached to PO via their NH2-terminal ends, while the COOH termini protrude from the ribosomal stalk (4,5). The COOH-terminal 17 amino acids of PO, P1, and P2 are almost identical except for conservative amino acid substitutions and are highly conserved between species (5-7). Selective elution of P1 and P2 (8,9) as well as antibody inhibition studies (10)(11)(12) suggest that the P proteins play an essential role(s) in protein synthesis. The P proteins are also of interest because approximately 15% of patients with systemic lupus erythematosus (SLE)' develop autoantibodies against these ribosomal proteins (13,14). The levels of anti-P autoantibodies fluctuate in patients with certain neuropsychiatric manifestations of the disease (15). The dominant epitope recognized by SLE autoantibodies has been localized to the common COOH terminus of the P proteins (16), and all sera tested bound to a synthetic peptide comprising the COOH-terminal22 amino acids (16). The COOH terminus of the P proteins contains a cluster of acidic amino acid residues surrounding potential phosphorylation sites at Ser-102 and Ser-105. These serines are within the sequence Glu-Glu-Ser-Glu-Glu-Ser-Glu-Glu-(Asp)-Glu, which represents a possible recognition site for casein kinase I1 (CK 11) (17,18). Since phosphorylation could potentially influence the antigenicity (19) and regulate the function (9,20) of the P proteins, we determined whether the P proteins are substrates of CK I1 and whether they were phosphorylated by CK I1 at their COOH termini in vitro and in uiuo.

MATERIALS AND METHODS
Preparation of Ribosomes and Isolation of PI and P2-Ribosomes were isolated from Hela and Ehrlich ascites cells, rabbit reticulocytes and Artemia salina cysts as described (13,21). In all cases, ribosomes were washed with buffer containing 50 mM Tris, pH 7.6, 500 mM KC1,5 mM MgC12 (T,,K500M5), and 250 mM sucrose. Where indicated, ribosomes were resuspended in TKM buffer containing 1% Triton X-100 and "stripped" from the endoplasmic reticulum and accessory proteins by ultracentrifugation through a discontinuous sucrose gradient (22). Total ribosomal protein concentrations were measured by the method of Bradford (23). P1 and P2 were selectively extracted from the ribosomes with T2,K7,M, and 50% ethanol (9) and were precipitated by the addition of 5 volumes of acetone.
Synthetic Peptides-Synthesis of the peptides comprising the carboxyl-terminal 22 amino acids of human P2 (C-22 peptide) and Artemia P2 (eL12) was by solid phase methods (16). The peptides were purified by reverse phase HPLC on a pBondapak C18 column.
The CK 11-specific peptide Arg-Arg-Arg-Glu-Glu-Glu-Thr-Glu-Glu- Glu (24)  Institute, Seattle, WA), was added at a dilution of 1:lOOO for reactions with exogenous CK 11. Following addition of the substrate (75 pg of ribosomal protein, 15 pg of hydrolyzed and partially dephosphorylated casein (Sigma), and/or varying concentrations of synthetic peptides), samples were incubated at 30 "C for 30 min unless indicated otherwise. Phosphorylation by the catalytic (C-) subunit of the CAMP-dependent protein kinase A (Sigma) was performed under the same conditions as those for CK

11.
Antisera-Human IgG anti-P autoantibodies were detected in the sera of patients with SLE by immunoblotting (26) and enzyme-linked immunosorbent assay using the C-22 synthetic peptide or recombinant P2 fusion protein antigens as described (27).
One-and Two-dimensional SDS-Polyacrylamide Gel Electrophoresis (PAGE) and Immunoblotting-One-dimensional SDS-PAGE was performed on 15% gels essentially as described by Laemmli (28). For two-dimensional gel electrophoresis (29) the first dimension gels contained the ampholytes pH 5-8 and 3-10 (Biolyte, Bio-Rad) in a 4:l ratio (13). Following SDS-PAGE and electrophoretic transfer to nitrocellulose paper, immunoblots were sequentially probed with primary (human SLE anti-P sera) and alkaline phosphatase-conjugated secondary antibodies and developed with nitroblue tetrazolium/5-11romo-4-chloro-3-indolyl phosphate. Dried gels and blots were exposed to Du Pont Cronex x-ray film with intensifying screens at -70 "C. Where appropriate, scanning laser densitometry of autoradiographs was performed.
Immunoprecipitation-IgG from SLE serum containing anti-P antibodies was adsorbed to protein A-Sepharose beads in phosphatebuffered saline (10 mM phosphate buffer, pH 7.4, 150 mM NaCl). The IgG-coated beads were washed and then incubated with ribosomes at room temperature for 2 h. The beads were washed with buffer containing 50 mM Tris, pH 7.5, 150 mM NaCl, 2 mM EDTA, 250 mM sucrose, 2.5% Triton X-100, and 1% SDS and eluted with Laemmli sample buffer.
Cell Culture and f"P]Orthophosphate Labeling-Ehrlich ascites cells (kindly provided by Dr. Francis M. Sirotnak, Memorial Sloan-Kettering Cancer Center, New York, NY) were maintained in Dulbecco's minimal essential/F12 medium supplemented with 10% fetal calf serum and 1% L-glutamine. Prior to labeling in uiuo, cells were suspended at 2 X 106/ml in phosphate-free medium for 30 min. The medium was removed and the cells were resuspended in 1 ml of the same medium containing 2 mCi of [:"P]orthophosphate. After 2 h, the labeling medium was replaced with 1 ml of Tl,Kl,Ml., buffer, pH 7.8, and the cells were disrupted by vortexing. The nuclei were pelleted by centrifugation at 8000 X g for 5 min, and the supernatant was used as cytoplasmic extract.
Tryptic Digestion and Two-dimensional Peptide Mapping-P1 and P2 phosphorylated in uitro were purified by preparative SDS-PAGE and electroelution (30). The phosphorylated (2-22 peptide was purified by reverse phase HPLC as described above. The P proteins phosphorylated in uiuo were isolated by immunoprecipitation from the cytoplasmic extract of Ehrlich ascites cells, SDS-PAGE, and electroelution (30). Pl/P2 and the C-22 peptide were digested with trypsin by standard procedures (31). The peptides were first separated by high voltage electrophoresis on cellulose thin layer plates at 800 V for 45 min in 1% NH,HCOa buffer, pH 8.9 (32), followed by ascending thin layer chromatography in the second dimension using pyridine/ n-butyl alcohol/acetic acid/H,O = 60:75:15:60. Tryptic phosphopeptides were located by autoradiography.

Phosphorylation of the P Proteins by Exogenous CK II-
Stripped intact ribosomes obtained from Artemia were incubated with exogenous CK I1 in the presence of [y-"'P]ATP.
The ribosomal proteins were then resolved by SDS-PAGE and blotted onto nitrocellulose paper. The autoradiograph of the blot showed incorporation of radiolabel into several bands ( Fig. lA, lune 1 ) . The 36-, 17.5-, and 16-kDa molecular mass phosphoproteins were assumed to be the Artemia P proteins PO, P1, and P2. When the same piece of nitrocellulose was probed with an anti-P serum, positive bands with molecular masses identical to the putative P proteins on the autoradiograph were observed (Fig. lA, lane 2). Similarly, immunoblotting following separation of proteins by two-dimensional gel electrophoresis revealed co-localization of the signals obtained by autoradiography and immunoblotting (Fig. 1, B1 and B2). To determine whether SLE anti-P serum could immunoprecipitate the phosphorylated proteins, Artemiu ribosomes were phosphorylated with exogenous CK I1 and incubated with either anti-P or normal serum. As shown in Fig. lC, three labeled proteins of 36, 17.5, and 16 kDa were precipitated by the anti-P serum (lane 1 ) but not by the normal serum (lune 2 ) . Taken together, these results indicate that the three ribosomal P proteins are in uitro substrates of exogenous CK 11.
T o determine whether the P proteins were specifically phosphorylated by CK 11, stripped ribosomes were incubated with either CK I1 or with the catalytic subunit of protein kinase A. CK I1 phosphorylated the three P proteins, but only minimally phosphorylated other ribosomal proteins (Fig. 2, lune 1 ). Protein kinase A, on the other hand, phosphorylated several other ribosomal proteins, but did not incorporate phosphate into the P proteins (Fig. 2, lane 2). This was verified by the failure of SLE anti-P serum to immunoprecipitate the phosphoproteins labeled by protein kinase A. These observations indicate that the ribosomal P proteins are not nonspecifically phosphorylated by multiple kinases.
Ribosomal Protein Kinase Actiuity-High salt-washed, stripped, Artemia ribosomes were incubated with or without exogenous CK I1 in the presence of [y-32P]ATP. Although weak labeling of the P proteins was observed in stripped ribosomes incubated without exogenous CK I1 (Fig. 3A, lane  2), the addition of exogenous CK I1 resulted in a significant increase in phosphorylation of PO, P1, and P2 (Fig. 3A, lane   1). When unstripped Ehrlich ascites ribosomes were incubated with [y-"PIATP alone (Fig. 3A, lune 3 ) or with [y-"PI ATP together with exogenous CK I1 (Fig. 3A, lune 4 ) , no difference in the level or pattern of phosphorylation of the P proteins was observed, suggesting that the ribosome-bound kinase maximally phosphorylated the P proteins in uitro. In order to determine whether the ribosome-associated kinase had the same properties as those described for purified CK I1 To determine whether the synthetic CK 11-specific peptide Arg-Arg-Arg-Glu-Glu-Glu-Thr-Glu-Glu-Glu (24) could compete with the P proteins for phosphorylation by the endogenous CK 11-like enzyme, ribosomes were incubated with [y-"2P]ATP in the presence or absence of the peptide competitor. Fig. 4 shows that 1 mM of the peptide almost completely (78%) inhibited phosphorylation of the P proteins by the associated kinase (lanes 1 and 5 ) . The inhibition was concentration-dependent, with 0.1 mM producing only Localization of the Site of Phosphorylation of P1 and P2-T o identify the location of the CK I1 phosphorylation site on the P proteins, unstripped ribosomes were phosphorylated by the endogenous CK 11-like enzyme in vitro. PI and P2 were ethanol-extracted from the ribosomes as described under "Materials and Methods" and the eluted phosphoproteins analyzed by SDS-PAGE and autoradiography. P1 and P2 accounted for greater than 95% of radiolabeled proteins extracted from the ribosomes (not shown). PI and P2 were digested by trypsin and the resulting phosphopeptides separated on thin layer plates by high voltage electrophoresis in the first dimension and chromatography in the second. As shown in Fig. 6A, two peptide(s) of very similar mobility were phosphorylated by the endogenous kinase. Similarly, when rihosome-bound P1 and P2 were phosphorylated by exogenous CK 11, extracted, and subjected to tryptic peptide mapping, a phosphopeptide of almost identical x/y coordinates to the upper peptide described above was observed (Fig. 6R). The upper peptide also comigrated with the phosphorylated and trypsin-digested C-22 synthetic peptide (Fig. 6C). These results indicate that CK I1 and the rihosome-bound kinase  phosphorylate a serine residue(s) located at the COOH terminus of the P proteins. Since two-dimensional PAGE immunoblots revealed two additional spots for 1'1 and P2 following incubation with CK I1 in vitro (not shown), we infer that both Ser-102 and Ser-105 were phosphorylated in Litre. To test whether the same site(s) was phosphorylated in Lliro, P1 and P2 were isolated from Ehrlich ascites cells that had been metabolically laheled with [ "?]orthophosphate. The peptide map of PI and P2 isolated from Ehrlich ascites cells laheled in vivo (Fig. 6 D ) was virtually identical to that observed for PI and P2 phosphorylated in vitro. A  from Ehrlich ascites cells labeled i n vivo (Fig. 7 A ) or i n uitro (Fig. 7 B ) comigrated (Fig. 7C). Since the same peptides were phosphorylated by exogenous CK I1 and the ribosome-bound kinase i n uitro and P proteins labeled i n uiuo, it seems highly likely that CK I1 or a very similar kinase phosphorylates the P proteins in uiuo.
To exclude the possiblity that phosphorylation of other sites on P1/P2 (Fig. 5 ) occurred, P1/P2 labeled i n uitro and i n uiuo were subjected to tryptic digestion and one-dimensional high voltage electrophoresis as described above, except that samples were applied to the center of the plate. Electrophoresis was performed for 15 min, and the phosphopeptides were located by autoradiography. No phosphopeptides other than those described above were detected.

DISCUSSION
The P proteins are three of approximately seven eukaryotic ribosomal proteins known to be phosphorylated i n vivo (36). In this report, we have shown that all three P proteins are i n uitro substrates of exogenously added CK 11, that the target site for CK I1 phosphorylation is located within the highly conserved COOH terminus, and that the same site is phosphorylated in uiuo. The P proteins were not promiscuous targets of protein kinases, since protein kinase A phosphorylated many ribosomal proteins but did not phosphorylate the P proteins i n uitro. Conversely, the P proteins were the major ribosomal substrates of CK I1 i n uitro. We also observed that ribosomes contained an endogenous kinase activity which phosphorylated the P proteins. The endogenous kinase showed a pattern of ribosomal protein phosphorylation very similar to exogenously added CK 11. Phosphorylation of the P proteins by the ribosome-bound kinase was inhibited by heparin and utilized both ATP and GTP as phosphate donors, properties characteristic of CK I1 (33,34). Furthermore, the CK II-specific peptide (24) competitively inhibited the endogenous kinase, strongly suggesting that it is CK I1 or a CK IIlike enzyme. Since the activity of the kinase responsible for phosphorylating the P proteins was considerably decreased following stripping of accessory proteins off the ribosomes, it seems likely that the kinase is a cytoplasmic protein that transiently binds to the ribosome, rather than an intrinsic ribosomal protein. CK I1 is known to be located in the cytoplasm as well as the nucleus (37) and to phosphorylate several inititation factors involved in protein synthesis as discussed below. The ability of the C-22 peptide to inhibit phosphorylation of the ribosomal P proteins i n uitro, the low K,,, of phosphorylation of C-22 by CK 11, as well as the failure to identify additional phosphopeptides by tryptic mapping indicates that the P proteins are phosphorylated at their COOH termini. This finding concurs with that of Amons et al. (38) who observed that Ser-98 of the Artemia P2 was exclusively phosphorylated i n uiuo. Stoichiometric studies on the C-22 synthetic peptide as well as two-dimensional SDS-PAGE immunoblots of P1/P2 suggested that both Ser-102 and Ser-105 of the mammalian sequence were phosphorylated by CK I1 i n uitro. Tryptic peptide mapping of PI, P2, and the C-22 peptide phosphorylated i n uitro, as well as P1 and P2 labeled i n uiuo, confirmed that the COOH termini of these proteins were the targets for phosphorylation by CK I1 and the ribosome-bound kinase. These experiments also suggest that CK 11, or a closely related enzyme, phosphorylates P1 and P2 at their COOH termini i n uiuo. The detection of two closely spaced, acidic phosphopeptides is most likely explained by incomplete tryptic digestion or a post-translational modification that differs between P1 and P2.
The high degree of conservation of the COOH termini among species and their surface exposure on the ribosomal stalk suggest that they are important for the function of the P proteins ( 5 ) . The functions of the P proteins have been investigated by selectively eluting P1 and P2 from the ribosome (9) and by observing the effect of anti-P antibodies on ribosomal function (10)(11)(12). Ribosomal cores stripped of P1 and P2 do not bind EF-1 (39) or EF-2 and lose EF-2-dependent GTPase activity (8,9). Similarly, human and rabbit polyclonal anti-P antibodies inhibit EF-2-dependent GTPase activity (9,20) and protein synthesis in general (11). Using mouse monoclonal anti-P antibodies, Uchiumi et al. (12) showed that a monoclonal antibody which recognized the COOH terminus on all three P proteins inhibited both EF-2mediated GTP hydrolysis and poly(U)-directed polyphenylalanine synthesis, whereas antibodies which were specific for either P1 or P2 did not. Since the inhibition of GTP hydrolysis and polyphenylalanine synthesis could be reversed by preincubating the monoclonal antibody with a synthetic peptide corresponding to the COOH terminus, the authors (12) concluded that the functional site of the P proteins was located in the COOH terminus. Several studies have suggested that phosphorylation of P1 and P2 influences binding of these proteins to the ribosome and/or ribosomal function (9,20). MacConnell and Kaplan (9) reported that EF-2-mediated G T P hydrolysis and polyphenylalanine synthesis were dependent upon the state of phosphorylation of P1 and P2. They were, however, unable to restore these functions by rephosphorylating P1 and P2 with protein kinase A (9). Since we were not able to phosphorylate the P proteins on intact ribosomes with protein kinase A and these proteins do not have protein kinase A sites predicted by the primary amino acid sequences (6,7,40), failure to restore function with protein kinase A does not seem surprising. Reconstitution studies with P1 and P2 phosphorylated by CK I1 should be of considerable interest.
Since the activity of CK I1 responds to stimulation of cells with epidermal growth factor, insulin-like growth factor, and insulin (41, 42), it has been suggested that the enzyme plays a role in cellular signal transduction. CK I1 is known to phosphorylate several initiation factors (eIF-2, eIF-3, and eIF-4) of protein synthesis (43), but it is unclear whether phosphorylation of these factors by CK I1 or other kinases has any influence on protein synthesis (37,43). In the nucleus, CK I1 phosphorylates DNA-topoisomerase 11, RNA-polymerase I1 as well as Myc (35) and the myb-encoded oncoprotein (Myb) (44). Transcription by RNA-polymerase I1 was initiated when the inactive, unphosphorylated form was phosphorylated by CK I1 at the highly conserved, acidic COOH terminus (45). Phosphorylation of Myb by CK I1 also appears to be functionally important, since the binding of Myb to DNA was reduced when Myb was phosphorylated by CK I1 and the CK I1 target site was deleted in cell lines whose transformation was ascribed to Myb activity (44). Whether growth factor induced stimulation of CK I1 results in the phosphorylation of the ribosomal P proteins and thereby affects protein synthesis is the subject of ongoing studies.
Anti-P autoantibodies are highly specific for the disease SLE (13). Epitope mapping by partial proteolysis revealed that the major epitope recognized by anti-P autoantibodies is located within the COOH-terminal 22 amino acids (16). The existence of a CK I1 phosphorylation site in the immunodominant epitope of the P proteins raises important questions regarding the antigenicity and immunogenicity of the unmodified and post-translationally modified forms of the P proteins. The role of phosphate groups in anti-P binding activity